Survey of openEHR storage implementations

Efficiently storing and retrieving archetype-based patient information can be a challenging task. This paper surveys current archetype-based system implementations in the world and in particular the different approaches that have been used to create 13606- or openEHR-based storage repositories. Data is reported from systems with a few records to millions of records, including both deployed systems in production and experimental systems. Worldwide 21 projects were found, 4 did not reply and 1 did not provide data. Many systems (n=11) base their storage on RDBMS, then often (n=6) with some XML data fields. Dedicated XML (n=3) and object-oriented (n=2) databases were other examples of storage used. Query formalisms used include SQL, AQL, XQuery and XPath. Service interfaces via SOAP (n=12) or REST (n=6) are common. Most systems support dynamic configuration using new/changed archetypes and templates dynamically without system restart. Some (n=7) systems use demographic archetypes. In addition to the built in DBMS indexing mechanisms, one project reports use of an additional inverted index to achieve improved performance.